Prognosis for glioblastoma multiforme (GBM), the most common primary brain malignancy, is extremely poor. Post-operative, concurrent chemo- and radiotherapy was recently shown to prolong survival but this intensification of therapy has led to the recognition of pseudoprogression (PsP) as a highly problematic confound to treatment monitoring and clinical trial design. In PsP a positive treatment response mimics progressive disease (PD) on conventional MRI. PsP complicates the differentiation of early progressive disease (ePD) from a positive treatment response affecting decisions about therapeutic intervention and potentially putting responsive patients at fatal risk of errant discontinuation of chemotherapy. Given the cost, potential morbidity, and sampling errors associated with biopsy or re-resection to establish a diagnosis of recurrence, clinically reliable non-invasive means of distinguishing PsP from ePD are critically needed. Further ambiguities arise with the application of anti-angiogenic therapy, which is thought to normalize leaky tumor vasculature to produce an almost immediate decrease in T1 enhancement (pseudo-response). This additionally complicates the assessment of disease progression and limits the utility of conventional MRI for the optimization of individualized treatment regimes. Dynamic susceptibility contrast (DSC)-PWI is the prevailing standard for clinical PWI and has shown promise in discriminating tumor progression from treatment effects. Unfortunately, current state-of-the-art DSC-PWI techniques have important technical limitations that critically impede their clinical utility including low spatial resolution, frequent geometric distortion and susceptibility related artifacts, especially in the post-operative brain, and estimation errors because of the confounding effects of contrast extravasation from leaky tumor vessels. Most of these are inherent to dynamic MRI because of the need for both high spatial and temporal resolution. Unfortunately, attempts to correct one of these deficiencies inevitably amplify others. Innovative approaches to accelerate MRI by an order of magnitude are necessary to establish DSC-PWI as a practical, relevant, and robust method in a clinical setting. The overarching aim of this research is to generate sensitive, specific, and clinically relevant imaging biomarkers of tumor progression and treatment response through the development of advanced PWI approaches. Our group pioneered several accelerated imaging techniques that exploit symbiosis of undersampled, non-Cartesian, data-acquisition trajectories with model-based and compressed-sensing-like reconstruction. We hypothesize that this synergy can also be exploited to provide the acceleration required to improve resolution and minimize artifacts inherent to standard DSC-PWI, as well as enable post-processing required to minimize extravasation effects. We aim to develop dynamic imaging acquisition and reconstruction methods that deliver high resolution, high accuracy DSC-PWI with minimized artifacts while providing high reliability. These methods will be applied in a pilot study of GBM patients undergoing post-operative MRI for monitoring of tumor progression. The project will leverage several unique resources and capabilities in broad MR imaging and Neuro-oncology research programs at UW-Madison. If successful, the techniques will not only change care and treatment development in brain tumor patients, but will also be useful for the study, diagnosis, and clinical management of other diseases including acute stroke, dementias, and neurodegenerative diseases.